Cerebrospinal fluid (CSF) augments metabolism and virulence expression factors in Acinetobacter baumannii

In a recent report by the Centers for Disease Control and Prevention (CDC), multidrug resistant (MDR) Acinetobacter baumannii is a pathogen described as an “urgent threat.” Infection with this bacterium manifests as different diseases such as community and nosocomial pneumonia, bloodstream infections, endocarditis, infections of the urinary tract, wound infections, burn infections, skin and soft tissue infections, and meningitis. In particular, nosocomial meningitis, an unwelcome complication of neurosurgery caused by extensively-drug resistant (XDR) A. baumannii, is extremely challenging to manage. Therefore, understanding how A. baumannii adapts to different host environments, such as cerebrospinal fluid (CSF) that may trigger changes in expression of virulence factors that are associated with the successful establishment and progress of this infection is necessary. The present in-vitro work describes, the genetic changes that occur during A. baumannii infiltration into CSF and displays A. baumannii’s expansive versatility to persist in a nutrient limited environment while enhancing several virulence factors to survive and persist. While a hypervirulent A. baumannii strain did not show changes in its transcriptome when incubated in the presence of CSF, a low-virulence isolate showed significant differences in gene expression and phenotypic traits. Exposure to 4% CSF caused increased expression of virulence factors such as fimbriae, pilins, and iron chelators, and other virulence determinants that was confirmed in various model systems. Furthermore, although CSF's presence did not enhance bacterial growth, an increase of expression of genes encoding transcription, translation, and the ATP synthesis machinery was observed. This work also explores A. baumannii’s response to an essential component, human serum albumin (HSA), within CSF to trigger the differential expression of genes associated with its pathoadaptibility in this environment.


Results and discussion
CSF enhances the expression of genes involved in transcription and translation machineries, ATP production, and specific metabolic pathways without increasing growth rate. To uncover the impact of specific host environments at the transcriptional level, two different A. baumannii strains, A118 (low pathogenicity and high antibiotic susceptibility) and AB5075 (hypervirulent and multi-drug resistant), were exposed to CSF [27][28][29] . Transcriptomic analysis of A. baumannii A118, using a fold-change cutoff of log 2 > 1 (with adjusted P-value < 0.05), showed 275 differentially expressed-genes (DEGs), 7.76% of the total genes in the A. baumannii A118 reference genome. However, statistically significant changes were not observed when A. baumannii AB5075, a hypervirulent and highly resistant strain, was exposed to CSF under the conditions tested. Previous observations have shown that A. baumannii's response to different stimuli correlates with each particular strain's degree of pathogenicity. Less pathogenic strains induced more changes in their phenotypic behavior to overcome the stressful environment and persist 30 .
The analysis of A. baumannii A118 DEGs revealed an increase in the expression of many genes involved in gene expression processes and energy production machineries (Table 1 and Supplementary Table S1). Notably, a large proportion of ribosomal protein genes are overexpressed upon exposure to CSF. Among the ribosomal protein associated genes, 47 out of 55 displayed a significant increase of expression of twofold or more. Coincidently, key translation genes such as those encoding elongation factors (EF) EF-G, EF-F and EF-P were also up-regulated. Concurrently, the main genes of the transcriptional machinery (RNA polymerase) were similarly overexpressed. The rpoB and rpoC genes, which code for the beta and beta' subunits of RNA polymerase (core of the transcription machinery), were overexpressed with a log 2 fold just below 1. However, the gene encoding the alpha subunit was also upregulated with a log 2 fold change of 1.48 (Table 1 and Supplementary Table S1).
In addition, genes important for energy production in the cell were also upregulated upon CSF exposure. The atpIBEFHAGCD locus, an operon encoding the FoF 1 -ATP synthase (the main ATP generator in the bacterial cell) displayed a threefold transcriptional increase in expression (Table 1 and Supplementary Table S1). These transcriptional responses suggest that when CSF is present in the environment, there is an increase in expression of transcription-and translation-related genes, as well as of FoF1-ATP synthase, the main ATP generator. Also, CSF exposure induces the transcription of specific metabolic routes in A. baumannii. In particular, several dehydrogenases of the tricarboxylic acid cycle intermediates as well as a citrate symporter were significantly overexpressed, together with two proline symporters (Supplementary Table S1). The transcriptomic data showed that type I glutamine synthetase (AbA118F_3228) and the proline symporter putP genes are upregulated by a log 2 fold change of 0.58 and 0.87 respectively.
Studies on Salmonella typhimurium showed that putP codes for a proline permease, an integral membrane protein, that is the primary transport protein when this amino acid is the only carbon or nitrogen source 31 . The transcriptomic analysis showed upregulation (log 2 fold change 0.62) of AbA118F_2664, a gene that encodes a CitMHS citrate-magnesium hydrogen complex symporter (Supplementary Table S1). Proteins of the CitMHS family transport citrate-Mg 2+ complex coupled with one proton per complex molecule 32 . Interestingly, increased citrate levels help survival of A. baumannii in certain conditions 26 . Thus, the net effect of CSF might be an increase in expression of this transporter, which would lead to higher citrate intracellular concentrations that may result in increased rate of growth.
While most of the DEGs were upregulated, only 66 showed a lower transcription level. Among the most downregulation genes we found those coding for transporters (e.g. Mg2 + transporters, AbA118F_2476 and AbA118F_2477) and catabolic proteins (such as AbA118F_1645 Alkaline Phosphatase and AbA118F_0711 Acyl-CoA dehydrogenase) (Table 1 and Supplementary Table S1).
Gene ontology (GO) analysis was next undertaken to identify molecular functions and biological pathways associated to A. baumannii's adaptive responses to CSF. Consistent with the above mentioned DEGs, GO www.nature.com/scientificreports/ enrichment analysis revealed a statistically significant overrepresentation of the GO categories ATP synthesis coupled proton transport, translation, tricarboxylic acid cycle, and aerobic respiration by 13.8-, 8.8-, 4.9-and 4.3-fold, respectively (adj. P-value < 0.05). Other studies found that exposure of A. baumannii strains to amikacin, imipenem, and meropenem was associated with increased expression of genes involved in the tricarboxylic acid cycle, biosynthesis of amino acids, purines, and pyrimidines, as well as the operons related to ATP, RNA, and protein synthesis 33 . Taken together, these results suggest that various stressful environments (low nutrient availability and antibiotic treatment) induce expression of genes associated to energy production, protein synthesis, and metabolism in A. baumannii. As a consequence of these transcriptomic changes, the bacterial cell adapts to survive and accelerate metabolism rive in these hostile conditions. The CSF-mediated upregulation of genes coding for the elements necessary for transcription, translation, expression and ATP synthesis was not accompanied by a decrease in generation time (Fig. 1). These results suggest that cells respond to CSF enhancing the expression of pathways that produce specific effects rather than increasing growth capacity.
There are numerous reports supporting the hypothesis that increasing the expression of enzymes involved in transcription, translation, and synthesis of ATP is correlated with faster growth rate [28][29][30][31][32] . However, these growth differences were not evident in either of the A. baumannii strains in the presence of CSF. An attractive notion to explain this observation is that the increase in gene expression capabilities is channeled toward the synthesis of cell components necessary for survival in the human body, e.g., adhesins and pilins.
The data described in this set of investigations indicate that certain modifications in the A. baumannii metabolism are uncoupled from growth rate. This is not anticipated because bacterial cells are characterized by allocating resources to maximize growth according to the needs for each environmental condition 34 . Our data also suggest that in depleted medium such as CSF, A. baumannii may be allocating all possible resources towards metabolism using an uncoupled metabolism to optimize its survival.
CSF affects the expression of A. baumannii virulence genes. We next observed that the addition of CSF to A. baumannii A118 cultures induced an increase in the expression of a set of genes that code for virulence-associated functions such as type IV pili, iron uptake systems, the type VI secretion system (T6SS), and poly-N-acetylglucosamine (PNAG) production.
Type IV pili participate in microbial adherence as well as motility (gliding or twitching). While A. baumannii lacks flagellum-mediated motility, twitching and surface-associated motility was demonstrated in several strains 35,36 . Numerous studies on twitching and surface-associated motility in A. baumannii A118 showed dependence on changes in light and temperature 27 as well as on the components of the growth media. In particular, addition of HSA resulted in increased motility and concomitantly upregulation of the cognate genes 26 .
Exposure of A. baumannii A118 to CSF produced an increase in the expression of pilW (log 2 fold change 1.22), pilJ (log 2 fold change 0.43), fimA (log 2 fold change 3.09), fimB (log 2 fold change 2.32), and the fimbrial protein precursor AbA118F_3133 (log 2 fold change 1.91) ( Fig. 2A). All of the type IV fimbriae genes have been experimentally shown to be associated with motility, cell adhesion, and biofilm formation 37,38 . In addition, our transcriptomic data showed reduced expression in the biofilm associated genes csuABCD, the two-component system response regulator bfmR, and the bap ortholog (biofilm-associated protein, see Supplementary Table S1). Significant differences were not observed in biofilm formation or motility in the presence of CSF (Fig. S1). This result is in contrast to our previous studies with 4% pleural fluid. The absence of changes in biofilm formation and motility can be explained by the different compositions of each fluid. Pleural fluid is clearly inducing more changes than CSF when comparing the number of genes affected by both fluids (1120 vs 275, respectively). Components such as neutrophils, lymphocytes, monocytes, proteins, reactive oxygen species, and neutralization agents are found in pleural fluid and could contribute to the different effect 22 .
The presence of CSF was also correlated with higher expression of twelve genes associated with the acinetobactin iron uptake system ( Fig. 2A and Table S1). These genes are part of the ferric-acinetobactin receptortranslocation machinery (bauABDE, bauA log 2 fold change 1.67), the acinetobactin biosynthesis (basBDFGJ, www.nature.com/scientificreports/ basD log 2 fold change 1.86) and export (barB) 39,40 ( Fig. 2A and Table S1). Besides their direct role in iron uptake in the iron starvation conditions found in the human host, the products of basD and basA are needed for A. baumannii to persist and cause apoptosis of human alveolar epithelial cells 41 . Bacterial iron uptake systems that are virulence factors are usually highly regulated and are induced under conditions of iron starvation. The transcriptomic data showed that, genes that code for functions in siderophore biosynthesis, export, and import are upregulated of in the presence of CSF. This finding adds another regulatory signal that enhances expression of acinetobactin iron uptake system. This increase in expression could be directly related to growth in the host or to biofilm formation, which is dependent on efficient iron uptake 42 . The structures of bacterial biofilms are usually dependent on polysaccharides such as poly-β-1,6-N-acetyl-dglucosamine (PGA) or cellulose. Previous studies showed that functional production of PGA in Escherichia coli depends on the products of four genes, pgaABCD. pgaC and pgaD are essential for biosynthesis, and pgaB, which specifies a N-deacetylase, together with pgaA are needed to export the polysaccharide from the periplasm to the extracellular milieu 43 . All four homologs were significantly upregulated when A. baumannii A118 was cultured in the presence of CSF (log 2 fold change of 1.63, 1.63, 1.56 and 1.08 for pgaA, pgaB, pgaC and pgaD, respectively) ( Fig. 2A and Table S1). As expected, Congo red staining showed that A. baumannii A118 cells cultured in the presence of CSF produced higher levels of PGA (Fig. 2B).
Another system known to be involved in A. baumannii's virulence is the T6SS 44 . The transcriptomic data showed that nine out of the 14 A. baumannii A118 T6SS genes were significantly upregulated in the presence of CSF. These genes tssABCDHIKLM, code for essential components of the T6SS ( Fig. 2A and Table S1).
Other genes of interest related with resistance and pathogenesis of A. baumannii were analyzed. We analyzed the genes involved with antibiotic resistance, quorum sensing, osmotic stress, DNA damage, outer membrane vesicle production, and capsule formation. While many were not found to be differentially expressed, a nonstatistically significant upregulation of carO*, which is involved in selective uptake of basic amino acids and also found to be related with carbapenem resistance, was observed 45 . A non-statistically decrease in expression of bla OXA-69 , β-lactamase found in A118 genome, was also observed. Only abaI, which codifies for the autoinducer synthesis protein 46 , was differentially expressed among the quorum sensing genes, while the rest where downregulated with the exception of fadD (See Fig. S2). In addition, we observed that the osmotic stress regulators, bet (a high-affinity choline uptake protein) and betI (a transcriptional regulator) were differentially upregulated by a log 2 fold change of -1.20 and -1.40, respectively. For genes associated with capsule formation (K-locus), pgm which encodes a phosphomannomutase was upregulated by 1.08 log 2 fold (P-value = 2.16 E-21). Regarding the SOS response associated genes, any of them was differentially expressed (See Fig. S2. This result differs from our previous work with pleural fluid exposure, where we found an overrepresentation of the expression of SOS response associated genes difference that can be explain by the presence of reactive oxygen species in pleural fluid. Lastly, among the genes related with outer-membrane vesicle (OMV) production, ompA was the only differentially expressed (Fig. S2). OmpA is known to have a cytotoxic effect and is considered a key virulence factor www.nature.com/scientificreports/ associated with bacterial biofilm formation, eukaryotic cell infection, resistance to antibiotics and also poses immunomodulatory effects 22,47 .

CSF enhances the release of A. baumannii's cytotoxic agents.
An initial assessment of the effect of CSF on A. baumannii virulence was determined using cytotoxicity assays. Bacteria-free conditioned medium (BFCM) obtained from A. baumannii A118 and AB5075 cultured in LB with or without CSF was added to human embryonic kidney cells (HEK-293), and the cells were inspected after 1 h. Figure 2C shows that BFCM samples obtained from CSF-containing A. baumannii AB5075 cultures were significantly more cytotoxic than BFCM from cultures that lacked CSF. This increase in cytotoxicity was observed at all tested concentrations, 1.5% BFCM (P-value = 0.006), 5% BFCM (P-value = 0.001), and 50% BFCM (P-value < 0.0001). Conversely, BFCM obtained from A. baumannii A118 cultures containing CSF showed an increased cytotoxic effect only at the highest concentration tested (50%) (P-value = 0.002). The results of these assays show that CSF induces the release of one or more cytotoxic substances by A. baumannii (Fig. 2C).

CSF-treatment increases A. baumannii virulence. The effect of CSF on A. baumannii's virulence was
tested using the Galleria mellonella model 41,48 . Infection with A. baumannii A118 cultured in LB plus 4% CSF resulted in increased mortality compared to the infection with cells cultured in LB (Fig. 2D). These results were consistent with the transcriptional changes in expression of virulence genes observed in vitro.

HSA contribution in A. baumannii pathoadaptation when exposed to CSF. Previous work
showed that the presence of pleural fluid (PF) is correlated with modifications of the expression of more than 1100 A. baumannii genes including many virulence factors involved in motility, biofilm formation, efflux, T6SS, fibrinolytic activity and capsule genes 22 , and with an increase in cytotoxicity and immune evasion 30 . The experiments shown in previous sections indicate that CSF produces effects similar to those observed with PF such as increase in cytotoxicity and changes in expression of virulence genes. Both fluids, PF and CSF, share as a component, HSA. Thus, we posited that this component may be responsible for the effects produced by both human fluids. To test this hypothesis, we compared levels of expression of the Type 1 fimbrial protein FimA, the iron transport protein BauD, and the Type IV secretion protein VgrG in A. baumannii A118 cells cultured in LB or LB supplemented with either CSF or HSA.
RNA-seq showed that bauD and vgrG, were up-regulated in both, CSF (log 2 fold-change = 2.5247 and 1.3611 respectively) and HSA (log 2 fold-change = 0.8046 and 1.1063 respectively) 26 . On the other hand, the fimA gene was up-regulated with a log 2 fold change of 3.0938 in the presence of LB supplemented with CSF but surprisingly, it was slightly down-regulated in medium supplemented with HSA (log 2 fold change of 1.6434) (non-statically significant considering a P-value < 0.05) 26 . While the results obtained with bauD and vgrG seemed to uphold the hypothesis, those produced when assessing expression of fimA did not seem to support that HSA is responsible for CSF-induced upregulation.
To confirm the results obtained by RNA-seq, another set of experiments was carried out measuring levels of expression by quantitative PCR using total RNA from A. baumannii A118 cells cultured in LB or LB supplemented with either CSF, HSA-depleted CSF (dCSF), or dCSF + 0.2% HSA (Fig. 3A-C). Expression of vgrG1 was www.nature.com/scientificreports/ slightly up-regulated in presence of 4% CSF (2.527-fold; P-value 0.4113) and dCSF (1.586-fold; P-value 0.9040). However, supplementation of LB with dCSF + 0.2% HSA, resulted in a robust increase of 6.065-fold (P-value 0.0152) and 3.825-fold (P-value 0.0234) with respect to LB or LB supplemented with dCSF, respectively (Fig. 3B). Expression of bauD was increased by 8.064-fold (P-value 0.0009) and 7.544-fold (P-value 0.0012) in cells growing in LB supplemented with CSF or dCSF + 0.2% HSA, respectively. Moreover, a 15.847-fold increase (P-value 0.0009) was observed in levels of expression of bauD in cells growing in LB supplemented with dCSF + 0.2% HSA with respect to those in cells growing in LB supplemented with dCSF (Fig. 3C). Conversely, while a slight increase in fimA expression was noted in cells growing in LB supplemented with CSF (1.057-fold, P-value 0.9190) or dCSF (1.238-fold, P-value 0.1349), there was a decrease of 1.107-fold (P-value 0.6526) when the medium was LB supplemented with dCSF + 0.2% HSA. A comparison between levels of expression in cells growing in LB supplemented with dCSF + 0.2% HSA or dCSF showed that the presence of HSA resulted in 1.370-fold decrease (P-value 0.0384) (Fig. 3A). These results confirmed that while HSA may be responsible for increasing expression of certain genes, it is not for others like fimA. In fact, it seems to induce a weak but consistent inhibition of expression of fimA. Future studies will need to carry out to determine if other components of CSF produce an increase in fimA expression. Bacterial cells express genes that code for factors that allow growth in the hostile environments they encounter upon invading the human body. The results shown in this section indicate that HSA is one of the signals that triggers the expression of several A. baumannii genes when the bacterial cells are in contact with HSA-containing fluids, PF or CSF. Furthermore, our previous studies showed that HSA also enhances DNA acquisition through modulation of natural competence-related gene expression and affects the expression of genes related to motility, efflux pumps, pathogenicity and antibiotic resistance, among others 23,26 . These characteristics are not unique to A. baumannii, various bacterial pathogens and protozoa 49-52 modify gene expression to adapt and thrive within the host utilizing HSA as one of the signals. For example, in Bordetella pertussis, the causative agent of the whooping cough, albumin combined with calcium induces an increase in production and release of the major toxin, adenylate cyclase toxin (ACT) 52 . Another example is the case of Pseudomonas aeruginosa, in which the presence of albumin is correlated with increased expression of iron-controlled genes (pvdS and regA) 53 . In summary, our observations, together with the evidence available from studies with other bacteria, suggest an important role of HSA as signal for expression of genes and systems essential for survival within the human body.

Conclusion
A. baumannii is one of many causative agents of nosocomial bacterial meningitis, an infection associated with high morbidity and mortality rates. During the infection, bacteria can be found in CSF. In fact, of the several methods available for diagnosis of bacterial meningitis, CSF culture is the most favored 54 . This study describes changes in expression of numerous genes when A. baumannii is exposed to CSF (Fig. 4). These genes code for a variety of proteins that participate in the gene expression machinery, energy production, motility, metabolism, survival, and virulence factors among others. Our results in combination with previous work suggest that HSA may be a contributor in signaling these transcriptomic responses (Fig. 4). HSA is one of the main components of CSF and is also present in blood and PF, all body fluids that trigger similar responses in bacteria. Utilizing HSA as the signal for gene expression of elements that facilitate progression of the infection is an intelligent strategy that permits bacteria to sense the presence of human environments. However, not all strains respond equally when HSA is present, slight differences were identified when comparing A. baumannii strains A118 and AB5075. These changes are correlated with differences in levels of pathogenicity and probably the kind of infections that are more commonly caused by each variant.
Another remarkable effect of HSA on A. baumannii is the augmentation of natural competency 26,55 . Traglia et al. proposed that a random coil stretch in the structure of HSA is responsible for increasing the ability of A. baumannii to take up DNA. Identifying factors that are similar between human host environments may be potential therapeutic target candidates. Considering the pleiotropic effects caused by the presence of HSA on A. baumannii, an alternative path to design therapeutic agents against this infection could be to identify compounds that interfere with the ability of A. baumannii to detect HSA. Compounds that interact with the HSA regions that are detected by A. baumannii could mask the presence of the protein impeding expression of the necessary systems for survival and progression of the infection.

Material and methods
Bacterial strains and human fluids. Two A. baumannii strains already used in previous studies [27][28][29] , exhibiting different degree of susceptibility and virulence were used. A. baumannii A118 strain is known to be susceptible to variety of antibiotics 28,56 and A. baumannii AB5075 possesses increased virulence and is resistant to many antibiotics 29 .
The Pooled Human Cerebrospinal Fluid (CSF) sample was acquired from Innovative Research, MI, which is a certified vendor that obtains human samples from Food and Drug Administration (FDA)-approved facilities. The samples were collected from normal healthy individual donors and pooled. Lysogeny Broth, LB, supplemented with 4% CSF was used for all CSF conditions. The following concentration was used since 4% of other human fluids has been used in previously studies 22,30,57 . RNA extraction and sequencing. A. baumannii colonies (A118 and AB5075) were suspended in LB with or without 4% CSF and incubated with agitation for 18 h at 37 °C. Overnight cultures were then diluted 1:10 in fresh LB broth and incubated with agitation for 7 h at 37 °C. RNA was extracted from each strain using the TRI REAGENT Kit (Molecular Research Center, Inc., Cincinnati, Ohio, USA) as previously described 26 . Total RNA extractions were performed in two biological replicates for each condition. www.nature.com/scientificreports/ RNA sequencing was outsourced to Novogene (Novogene Corporation, CA) for mRNA-seq analysis, which includes rRNA depletion, library preparation following the protocols of the NEBNext Ultra II Directional RNA Library Prep Kit for Illumina (New England Biolabs) (New England Biolabs) and HiSeq 2500 paired-end 150 bp sequencing.
RNA-seq data analysis. RNA-seq reads (GEO accession No GSE153967) corresponding to A. baumannii strain A118 and AB5075 exposed to LB or LB plus 4% CSF were analyzed as follows. Trimming of low-quality bases at the ends of the reads to a minimum length of 100 bp and removal of Illumina adaptor sequences was performed using Trimmomatic 58 . FastQC (www.bioin forma tics.babra ham.ac.uk/proje cts/fastq c/) was used to assess the quality of the reads before and after trimming. Burrows-Wheeler Alignment software (BWA) was used to align the RNA-seq reads to sequences of the whole genome shotgun sequencing project of strain Acinetobacter baumannii A118F (DDBJ/ENA/GenBank accession VCCO01000000). FeatureCounts was used to calculate the read counts per gene, and differential expression analysis was performed using DEseq2 59,60 . Principal component analysis (PCA) and gene expression heat map with clustering dendrograms of the RNA-seq data analysis of LB and CSF treatments are shown in Supplementary Fig. S3. Features exhibiting FDR < 0.05 and log 2 fold change > 1 were considered statistically significant.
Gene ontology (GO) analysis. GO terms were retrieved from UniProt for the best BLASTx hits to A. baumannii A118F genes. Using GO.db Bioconductor annotation data package in R language, GO terms and ancestor terms were assigned for all DEGs from this study. GO enrichment analysis was performed using custom-made scripts as described previously 61 . The enrichment factor was estimated as the ratio between the proportions of genes associated with a particular GO category present in the dataset under analysis, relative to the proportion of the number of genes in this category in the whole genome. P-values were calculated using the Fisher Exact Test and adjusted by the Benjamini-Hochberg method.
Growth curves. Growth curves of both strains, A118 and AB5075, were conducted on 96-well plates. Overnight cultures were subcultured 1:50 in LB or LB + 4% CSF and incubated for 18 h at 37 °C with medium shaking. OD 600 nm was measured every 20 min using a Synergy 2 multi-mode plate reader (BioTek, Winooski, VT, USA) and Gen5 microplate reader software (BioTek). To study the effect of different carbon sources, both strains were culture overnight under different condition (LB broth and LB broth + 4% CSF). www.nature.com/scientificreports/ Motility and biofilm assays. Motility and biofilms assays were performed as previously described 26 . A118 and AB5075 cells were cultured in LB broth with or without 4% CSF. Bacterial cells were incubated with agitation for 18 h at 37 °C. Experiments were performed in triplicate, with at least three technical replicates per biological replicate.
Cytotoxicity assays. In a Nunclon Delta Surface opaque 96-well microplate (ThermoScientific), we added colorless DMEM, 4% CSF, and A118 or AB5075 BFCM diluted in LB broth to make 50 μL of BFCM at final concentrations of 1.5%, 5%, and 50%. An additional 50 μL of ATCC HEK-293 cells at a concentration of 1 × 10 6 cells/mL in colorless DMEM were suspended in the well and intoxicated for 1 h at 37 °C, 5% CO 2 . CellTiter-Glo Reagent (100 μl) was added to each experimental and standard curve well and then placed on an orbital shaker for 2 min. Following mixing, plates were incubated at room temperature for 10 min to stabilize the luminescent signal. The viability of HEK cells was measured at room temperature using the "all" luminescence function of SpectraMax M3. Galleria mellonella infection model. To assess the virulence of A. baumannii with and without CSF in vivo, the G. mellonella insect model of infection was used 62 . Larvae weighing between 200 and 400 mg were maintained on wood chips in the dark at 4 °C. A. baumannii A118 was grown overnight in either LB or LB with 4% CSF. An equivalent of 1.0 OD 600 unit of overnight culture was pelleted and resuspended in 1 mL of cold sterile 20 mM phosphate buffered saline, pH 7.4 (PBS). The cells were further diluted 1:10 in sterile PBS and used for injections. A Hamilton syringe was used to inject 5 μL of the diluted bacterial suspension via the left proleg of each larva. A control group of untreated larvae was used to assess overall larval viability for the duration of the assay. One hundred G. mellonella larvae were used in each condition and incubated at (37 °C) in a sterile Petri dish for 24 h intervals for 48 h total. Larvae viability was monitored by observing response to gentle prodding with a glass rod; those with no response were considered dead. Four replicates with 100 larvae per Petri dish were performed for each condition.

PNAG assays.
To study extracellular matrix (ECM) production, microcolony biofilm was used as model system. 5 μl of overnight cultures of A118 cultured in LB broth and LB broth + 4% CSF were inoculated on LB agar and supplemented with Congo red as previously described 63 . Plates were incubated at 28 °C in static incubator for up to 48hs. Results were recorded at 24 h with a Plugable USB 2.0 Digital Microscope.
HSA depletion. HSA was depleted from CSF by placing 1 mL of CSF into a 30 kDa Amicon Ultra Centrifugal Filter (Millipore, Temecula, CA, United States) and the solution was centrifuged at 20,000 × g for 10 min.

RT-qPCR.
Previously extracted and DNase-treated RNA from A. baumannii strain A118 grown in LB, 4% CSF, 4% depleted CSF and 4% depleted CSF + 0.2% HSA, were synthesized to cDNA using the manufacturer protocol provided within the iScript Reverse Transcription Supermix for qPCR (Bio-Rad, Hercules, CA, United States). The cDNA concentrations were measured with a DeNovix DS-11 + spectrophotometer; each sample was then diluted to a concentration of 50 ng/μl. qPCR was conducted using the iQ SYBR Green Supermix through the manufacturer's instructions. At least three biological replicates of cDNA were used and were run in quadruplet. All samples were then run on the CFX96 Touch Real-Time PCR Detection System (Bio-Rad, Hercules, CA, United States). The transcript levels of each sample were normalized to the recA rRNA transcript levels for each cDNA sample. The relative quantification of gene expression was performed using the comparative threshold method 2 −ΔΔCt . The ratios obtained after normalization were expressed as folds of change compared with cDNA samples isolated from bacteria cultures on LB. Asterisks indicate significant differences as determined by ANOVA followed by Tukey's multiple comparison test (P < 0.05), using GraphPad Prism (GraphPad software, San Diego, CA, United States).

Statistical analysis.
All experiments were performed at least in technical and biological triplicate. Data was expressed as means ± standard deviation. Statistical analysis using Mann-Whitney test or ANOVA followed by Tukey's multiple comparison test were performed using GraphPad Prism (GraphPad software, San Diego, CA, USA), and a P-value < 0.05 was considered statistically significant.
All procedures performed in this study were in accordance with the CSUF Institutional Biosafety Committee Approval plan (DBH117-01) and are in compliance with the NIH, CDC, OSHA and other environmental and occupational regulations.